outline - cibmtr
TRANSCRIPT
1
Comorbidities before Allogeneic
Hematopoietic Cell Transplantation (HCT)
The HCT-specific Comorbidity Index (HCT-CI)
Mohamed Sorror, M.D., M.Sc.FHCRC
Seattle, WA
Outline• Pretransplant Essential data
• Why comorbidities are important?– For patients with cancer– For patients given allogeneic HCT
• What is the HCT-CI?
• How to collect comorbidities per the HCT-CI?
• How the HCT-CI scores could be utilized?– Outcome prediction– Comparing trials at different institutions– Causes of death– Other pretransplant risk factors
Pre-Transplant Essential Data
2
Why comorbidities are important?
For patients with cancer
BackgroundComorbidity
• Any distinct additional clinical entity that has existed or may occur during the clinical course of a patient with an index disease. (Feinstein. J Chronic Dis. 1970; 23:455)
• Relevant in the prognosis of cancer patients.
• Physiological burden of chronic disease and its interaction with cancer and cancer treatment
• Increased severity of comorbidities leads to increased risks of toxicities to specific therapies
Comorbidities and Cancer
Clinical impacts• Prognosis• Quality of care
– Choice of therapy– Tolerability to therapy– Mortality
• Quality of life (QOL)
Statistical impact• Confounder• Effect modifier• Predictor of study outcome• One comprehensive measure
de Groot V. J. Clin. Epidemiol., 2003
3
Comorbidities and CancerSolid Tumors ± Lymphoma
• Major predictors of quality of life– Age– Comorbidities (scores or numbers)– Cancer site– ± Symptom severity
• Age did not constitute a difference (<45, 45-65, >65)• Comorbidities caused age-related differences
Greimel ER. British J Cancer, 1997
Comorbidities and Hematological Malignancies
• Impact survival
– Death from comorbidities
– Contraindications to specific therapy
– Reduction of specific-therapy dose
– Treatment-related complicationsLymphoma: van Spronsen DL. Euro. J Cancer, 2005
Leukemia: Pinto A. Critical Reviews, 2001
LymphomaLeukemia
Why comorbidities are important?
For patients given allogeneic HCT
4
• Myeloablative conditioning regimens:
– High-dose chemotherapy and/or radiotherapy
– Potentially curative treatment
– Relatively high NRM
– Young healthy patients with limited prior treatment history
• Nonmyeloablative or reduced-intensity conditioning regimens:
– Milder regimen-related toxicity and mortality
– Enroll older patients
– Enroll more patients with comorbidities
Allogeneic HCT for hematological malignancies
• Available literature focuses on studying individual
comorbidities
• No data on the impact of increasing number and severity of
comorbidities on HCT outcomes
• No comorbidity objective measures are available to
determine which patients:– Could tolerate myeloablative conditioning
– Would benefit from nonmyeloablative conditioning
– Would not benefit from either kind of conditioning
Allogeneic HCT for hematological malignancies
Initial experiences with comorbidity indices in allogeneic
HCT
The Charlson Comorbidity Index (CCI)
5
The CCI
• Developed from– Number and severity of comorbid diseases– An inception cohort of 604 medical patients– Admitted for a 1 month period at NY hospital– One-year follow up data
• Comorbidity weighted index (training population):– Relative risks of each comorbidity for 1-year mortality– Adjusted for
• All other comorbidities• Illness severity• Reason for admission
– Employed as weights for different comorbidities
Charlson et al, J Chronic Dis., 1987;104:961-968
CCI scores and grade IV toxicity: URD
89%
66%
Days after HCT
% o
f pat
ient
s
Nonmyeloablative (n=60)
p=0.06
Myeloablative (n=74)
p=0.1
Score 0Score 1-2Score ≥3
Days after HCTSorror ML et al, Blood. 2004, 104(4): 961-8.
Seattle experience
55%38%21%
28%
67%p=0.03
CCI scores and NRM: URD
14%
19%36%
% o
f pat
ient
s
Months after HCT Months after HCT
p=0.1
Nonmyeloablative (n=60) Myeloablative (n=74)
Seattle experience
Sorror ML et al, Blood. 2004, 104(4): 961-8.
Score 0Score 1-2Score ≥3
6
63%
22%
CCI scores and survival: URD
Months after HCT
57%
47%
9%
Months after HCT
% o
f pat
ient
s
p=0.02 p=0.03
Score 0Score 1-2Score ≥3
Nonmyeloablative (n=60) Myeloablative (n=74)
Seattle experience
Sorror ML et al, Blood. 2004, 104(4): 961-8.
0.02264% NRM Day360
0.02143% NRM Day100
0.00043876% EFS @ 2yrs.
0.0015383% OS @ 2yrs.
PScores >2 (n = 50)
Scores 0-2 (n = 28)
Outcomes for all patients
CCI scores and outcomes: AML/MDS
Houston experience
Giralt S et al, Tandem BMT Meeting 2004
The HCT-CI
Sorror et al, Blood. 2005, 106(8): 2912-9
7
• Some CCI comorbidities were rarely identified among HCT patients– Pulmonary and hepatic comorbidities
• Some comorbidities were not represented in the CCI– Infections and depression
• Lack of sensitivity– Scores ≥1 in 35% of patients
• Particularly low in myeloablative patients (12%)
Limitations of the CCI for HCT
Aims• Better define comorbidities using laboratory data
– Pulmonary, hepatic, cardiac, and renal comorbidities
• Investigate additional comorbidities among HCT patients
– All new comorbidities
• Establish comorbidity scores suited for HCT
– Other comorbidities as originally defined
Design: 1055 patients
• Nonmyeloablative = 294• Myeloablative = 761
• Transplanted between– 1997- 2003 for related recipients– 2000-2003 for unrelated recipients
• Malignant or non-malignant hematological diseases
• Database pre-HCT lab values– Bilirubin, AST, ALT, creatinine, EF, DLco, FEV1
• Retrospective chart review
8
Developing the new HCT-CI
• All patients were randomly divided into two populations
– Training set (n=708)
– Validation set (n=347)
Refining some comorbidity definitions
DLco/ FEV1 >80%Dyspnea grade IIMild
EF ≤50%Heart failure & myocardial infarction
Cardiac
DLco/ FEV1 ≤65%Dyspnea grade IVSevere
DLco/ FEV1 66-80%Dyspnea grade IIIModerate
Pulmonary:
Added definitionHCT-CI
Old definition CCI
Comorbidity
Refining some comorbidity definitions
Creatinine >1.2-2 mg/dlCreatinine 2-3 mg/dlMild
Hepatitis, bilirubin (1.5 ULN), or AST/ALT (2.5 ULN)
Hepatitis or cirrhosisMild
Creatinine >2mg/dlCreatinine >3 mg/dl, renal dialysis/transplant
Moderate/severe
Renal:
Cirrhosis, bilirubin (>1.5 ULN), or AST/ALT (>2.5 ULN)
Portal hypertension ±bleeding varices
Moderate/severe
Hepatic:
Added definitionHCT-CI
Old definition CCI
Comorbidity
9
The added new comorbidities
4%
2%
9%
1%
1%
4%
1%
Prevalence
Obesity
Depression/anxiety
Infection
Osteoprosis
Osteoarthritis
Migraine/headache
Bleeding
New comorbidity
Development of the scoresTraining set
• 28 comorbidities• Calculation of unadjusted HR of each comorbidity for
– NRMAt 2-years
• Adjustment of HRs for– All other comorbidities– Disease risk– Type of conditioning– Age
• Adjusted HR were employed as weights:– HR ≤1.2 = score 0– HR 1.3-2.0 = score 1– HR 2.1-3.0 = score 2– HR 3.1-3.9 = score 3
New scores and prediction of NRM
Training set (n = 708)
HR*NRM,%
# Pts,%
Scores
6.934311≥ 4
6.0941173
3.4827172
1.6614171
19380
*Adjusted for age, disease risk and type of conditioning
Validation set (n = 346)
HR*NRM,%
# Pts,%
3.054013
3.954115
1.261917
1.572218
11438
10
Per
cent
sur
viva
l
CCI
Per
cent
NR
M
The validation set
Score 0 (87%)Score 1 (10%)Score ≥2 (3%)
Score 0 (38%)Score 1-2 (34%)Score ≥3 (28%)
HCT-CI
Months after HCT
HCT-CI
Scoring comorbidities
Cardiovascular• Arrhythmia:
– Atrial fibrillation (AF)– Atrial flutter– Ventricular arrhythmias (Tachycardia or fibrillation)– Sick sinus syndrome
• Cardiac problems:– Coronary artery disease– Myocardial infarction– Congestive heart failure– Ejection fraction (EF) ≤50%
• Valvular disease– Any proven valve stenosis or malfunction with the exception of
asymptomatic mitral valve prolapse– Prosthetic aortic or mitral valves
11
Gastrointestinal
• Inflammatory bowel disease– Ulcerative colitis– Crohn’s disease
• Peptic ulcer– Previously required treatment– Previously bled from ulcer
Endocrine
• Diabetes
– Type I – Type II
• requiring treatment with oral hypoglycemic drugs or insulin
Neurology
• Cerebro-vascular disease
– History of transient ischemic attacks
– History of a cerebro-vascular accident
12
• Obesity– Patients with body mass index of >35
(weight in kg/ height x height in m)
• Infection– Documented or suspected and requiring
treatment before, during, and after start of conditioning regimen
• Psychiatric disturbances– Depression– Anxiety
• Previously diagnosed and receiving specific treatment
• Diagnosed and started treatment at the time of HCT
• Renal– Serum creatinine >2 mg/dl– On dialysis– Had prior renal transplantation
• Preceding solid malignancy– Requiring treatment– Excluding non-melanoma skin cancer
13
• Rheumatologic– Systemic lupus erythmatosis (SLE)
– Rheumatoid arthritis (RA)
– Polymyositis
– Mixed connective tissue disease
– Polymyalgia rheumatica
Hepatic• Mild
– Chronic hepatitis– Bilirubin >upper limit of normal (ULN)-1.5 x
ULN– AST or ALT >ULN-2.5 x ULN
Or• Moderate-severe
– Cirrhosis or fibrosis proved by liver biopsy– Bilirubin >1.5 x ULN– AST or ALT >2.5 x ULN
Pulmonary• Moderate
– Diffusion capacity of CO (DLco) 80%-66%– FEV1 80%-66%– Shortness of breath on exertion
Or• Severe
– DLco ≤65%– FEV1 ≤65%– Shortness of breath at rest– Requiring supplemental oxygen
14
Current and future applications of the HCT-CI
• Predict HCT outcomes
• Determine how patients tolerate different conditioning regimens
Correlation with Conditioning Intensity
Patients diagnosed with MDS or AML
Sorror ML et al, ASH 2005
15
0
10
20
30
40
50
60
0 1 2 3 ≥4
NonablativeAblative
HCT-CI scores
% o
f pat
ient
s
Nonmyeloablative
Myeloablative
HCT-CI scores
Multivariate analysis Risk factors for NRM
0.031.59Recipient CMV positive
0.0022.00High-risk disease
0.031.92myeloablative conditioning
0.011.75Marrow
0.011.82Recipient age ≥50 years
0.0051.80Unrelated donor
<0.00013.32HCT-CI scores of ≥2
PHRRisk factors
nonmyeloablative
myeloablative
% N
RM
Adjusted*P = 0.006
*Adjusted for age, diagnosis, stem cell source, donor type, prior HCT, and CMV serostatus
Years after HCT
Patients with high-risk MDS/AML and HCT-CI scores of ≥2
16
nonmyeloablative
myeloablative% O
vera
ll su
rviv
al
Years after HCT
Adjusted*P = 0.01
*Adjusted for age, diagnosis, stem cell source, donor type, prior HCT, and CMV serostatus
Patients with high-risk MDS/AML and HCT-CI scores of ≥2
Summary• Patients with high-risk MDS or AML and HCT-CI
scores of ≥2 receiving nonablative conditioning: – Less NRM– Survival benefit
• More data are needed for patients with:– Low-risk MDS or AML– Low HCT-CI scores
• HCT-CI should be considered among other risk factors:
– Prospective trials– Patient counseling
Lymphoid malignancies
Correlation with Conditioning Intensity
Sorror ML et al, ASH 2006
17
%%Diagnoses
7953NHL
1527CLL
620HD
3445URD
6999G-PBMC
2 (0 - 8)3 (0 - 10)Median (range) # of prior regimens4052Median age, years
1045Prior HCT
48/353/13Refractory/untreated relapse
Ablative(n=68)
Nonablative(n=152)
Characteristics
*HR: 0.95, p = 0.91Survival*NRM*
%
HCT-CI score 0: Adjusted outcomes
*Adjusted for age, stem cell source, diagnoses, CMV sero-status, prior regimens, donor type, and disease chemo-sensitivity
*HR: 0.92, p = 0.93
MyeloablativeNonmyeloablative
Nonmyeloablative (adjusted)
Years after HCT
*HR: 0.33, p=0.0007Survival*
Years after HCT*Adjusted for age, stem cell source, diagnoses, CMV sero-status, prior regimens, donor type, and disease chemo-sensitivity
HCT-CI score ≥1: Adjusted outcomes
%
*HR: 0.19, p<0.0001NRM*
MyeloablativeNonmyeloablative
Nonmyeloablative (adjusted)
18
Summary and future directions
• Patients with no comorbidities: At 3-years– NRM of 15%-18% regardless of conditioning
Lymphoid malignancies
• Patients with comorbidities receiving nonablative compared to ablative conditioning:
– Less NRM– Better survival
• Prospective randomized studies for patients:– NO comorbidities– Age ≤60 years
Comparing patients at different institutions
MD Anderson Cancer Center
Patients with AML in 1st complete remission
Sorror ML et al, Tandem BMT 2006
5.14.8Interval from Dx to HCT, months
Mismatched related
15%28%Unrelated
7%4%
Matched related
39 (19 – 67)41 (19 – 75)Age, median (range) years
42%35%G-PBMC
78%68%Donors
MDACC(n=67)
FHCRC(n=177)
CharacteristicsAML in first complete remission
19
Non TBI-based
TBI-based
2444
4345Myeloablative
1010Nonablative§
222Reduced intensity*
MDACC(n=67), %
FHCRC(n=177), %
Conditioning regimens
*Includes fludarabine/alkylating agent-based regimens§Includes 2 Gy TBI-based regimen or fludarabine/Ara-C/idarubucin
0
10
20
30
40
50
60
0 1 2 3 ≥4
FHCRCMDACC
HCT-CI scores
% o
f pat
ient
s
FHCRCMDACC
HCT-CI scores
0
10
20
30
40
50
60
Pulmonary Hepatic Malignancy Cardiac Obesity Infect ion DM Depression Rheumatology
FHCRCMDACC
Individual comorbidities
Lung CancerLiver Obesity InfectionCardiac
% o
f pat
ient
s
DM Psych Rheum
FHCRCMDACC
20
2-year NRM stratified by HCT-CI scores
Years after HCT
Per
cent
NR
M
Score 0 Score 1-2 Score ≥3
FHCRC MDACC
7
19
37
721
27
2-year survival stratified by HCT-CI scores
Years after HCT
Per
cent
sur
viva
l
FHCRC MDACCScore ≥3Score 1-2Score 0
High Early Mortality
Palliative Care Discussion Programs
21
Group 1
Conditioning
Group 2
Myeloablative Nonmyeloablative
Relapse-risk High* High*
HCT-CI scores ≥3 ≥6
*All except AML in 1st CR, CML in 1st chronic phase, MDS-RA or RARS
0 1 2 3 40
25
50
75
100
MyeloablativeHCT-CI scores ≥3
0 1 2 3 40
25
50
75
100
NonmyeloablativeHCT-CI scores ≥6
Years after HCT
%
Patients with High-risk for relapse
HCT-CI and acute GVHD
Sorror ML et al, Tandem BMT 2006
• Correlation with causes of death
22
Patients
• Hematological malignancies
• 1993-2002
• myeloablative conditioning
• CSP/MTX GVHD prophylaxis
17%
83%
66%
34%
BU/CY
CY/TBI
Conditioning
3842Median age, years
56%53%≥1HCT-CI scores
77%67%MarrowStem cell source
58%54%LowDisease-risk
Unrelated(n=249)
Related(n=709)Characteristics
Factors analyzed
• Recipient age
• Donor age
• Recipient/donor sex
• Type of donor
• TBI versus no TBI
• Dose of TBI
• HCT-CI
• Disease risk
• Year of HCT
• Marrow vs. G-PBMC
• Full vs. reduced dose:
– CSP
– MTX
23
< 0.00011.85UnrelatedDonor
0.0011.42>20Recipient age
0.051.95< 80%CSP dose reduction
<0.00011.71YesTBI
0.0081.63≥5HCT-CI scores
PHRRisk factors
Multivariate analysisRisk factors for grades II-IV acute GVHD
< .00012.60UnrelatedDonor
0.021.96< 80%MTX dose reduction
0.0081.95< 80%CSP dose reduction
0.0081.55HighDisease-risk
< 0.00013.05≥5HCT-CI scores
PHRRisk factors
Multivariate analysisRisk factors for grades III-IV acute GVHD
<0.000173%34%CSP dose reduction
0.0002
P
46%68%HCT-CI scores of 0
≥ 20 years< 20 years
Recipient age
Correlation between age, comorbidity and CSP dose reduction
24
Grades III-IV acute GVHDHCT-CI scores
Days after HCT
Pro
babi
lity
if G
VH
D
Score 0
Score 1-3
Score 4
Score ≥5
Related
Unrelated
Summary
• The age related reductions of CSP dosing were likely due to increasing comorbidities with increasing age.
• Patients with high comorbidity scores or advanced disease should be stratified in future clinical trials for GVHD prophylaxis.
HCT-CI and Performance Status
Nonmyeloablative Conditioning
• Interaction with other pretransplant risk factors
Sorror ML et al, ASH 2006
25
Correlation between HCT-CI and KPS scores
r = - 0.18
HC
T-C
I sco
res
KPS scores
NRM
Years after HCT
Per
cent
NR
M
HCT-CI scores 0-2
HCT-CI scores ≥3
KPS >80%
KPS ≤80%34
29
1622
HR: 2.91
P<0.0001
HR: 1.64
P = 0.04
Years after HCT
Per
cent
sur
viva
l
Survival
HCT-CI scores 0-2
HCT-CI scores ≥3
KPS >80%
KPS ≤80%37
44
6557
HR: 2.46
P<0.0001
HR: 1.58
P = 0.01
26
Consolidated HCT-CI and KPS
II
I
≤80%
>80%
≤80%
>80%
KPSHCT-CI
21≥3High
25≥3
160-2Intermediate
380-2Low
Patients (n=341)
%
DefinitionsRisk Groups
Consolidated HCT-CI and KPSImpact on Survival
% o
f pat
ient
s
Years after HCT
Low
Intermediate I
High
Intermediate II
P < 0.0001
68
58
41
32
Conclusion: HCT-CI
• Higher performance
• Compare trial results at different institutions
• Stratify patients– Nonmyeloablative versus myeloablative
• Prognostic
• Consolidated with Performance Status
• One measure for different comorbidities
• Impact on GVHD prophylaxis regimens
27
Future Aims
• Multi-institutional validation
• Inter-rater reliability
• Post-HCT toxicities and quality of life
• Causes of death
• Simplifying assessment– Different methods– Education program
• Interaction with aging